Livelock is a failure mode where a system is not frozen, but it still makes no useful progress. Threads, processes, or kernel routines keep waking up, retrying, yielding, or rechecking state, yet they never reach completion. Unlike a deadlock, the system may stay busy and responsive at the surface while useful work stalls underneath.
In cybersecurity, livelock matters because it can become a denial-of-service condition. Attackers may trigger edge cases, contention, or repeated retries so a service burns CPU without serving requests. In kernel code, bad lock handling, interrupt storms, or overly aggressive retry loops can produce this behavior. Defenders look for symptoms such as high CPU usage, repeated log messages, timeouts, and watchdog alerts. Fixes usually involve better lock ordering, backoff, timeouts, and limits on retry loops so the system can recover instead of spinning forever.